Flitto Unveils AI Interpretation Brand Vogl, With Personalization at Its Core

Flitto has consolidated its AI interpretation and translation product lineup under a single brand and set out to expand its business around personalization. On the 30th, Flitto held its "Media Insight Day" at the Grand InterContinental Seoul Parnas in Seoul and unveiled Vogl, the integrated brand for its AI interpretation and translation solutions, short for "Voice of Global Languages."

Lee Jeong-su, CEO of Flitto, speaks at the "Media Insight Day" held on the 30th. In his presentation, Lee emphasized personalization.

Flitto's personalization approach involves continuously learning a user's voice, as well as the words and expressions they use frequently. Through this, the company aims to improve interpretation and translation accuracy for each user.

Flitto said it is placing personalization at the center because, in the interpretation and translation process, it is difficult to recognize proper nouns such as people's names and company names, and it is also difficult to recognize specialized terms by industry. In addition, differences in pronunciation from person to person were cited as another reason personalization is needed.

Flitto explained that such recognition failures affect translation quality. Accordingly, the company has put personalization based on each user's voice, vocabulary, and expressions at the center of its business expansion strategy.

Flitto will apply its personalization function to Vogl's various solutions. Vogl consists of a personal AI interpretation app, Vogl Stage for conferences, seminars, and education, Vogl Zone for installation in airports, banks, and stores, a dedicated solution for Cisco Webex, and API.

Lee said that pronunciation differs from person to person and that proper nouns also vary by person. He added that interpretation requires an understanding of the individual, and that Vogl's strength lies in understanding the user.

When AI fails to recognize a word or expression, Vogl asks the user to confirm it and then relearns the confirmed word or expression. Voice and terminology information accumulated in the app and on the web is also linked to and used in other interpretation and translation environments such as Vogl Stage.

Behind this implementation of personalization in Vogl is Flitto's data business capabilities. Flitto began with multilingual text and expanded into multimodal data such as voice, images, and video, as well as data for inference and evaluation. Recently, it has broadened its scope to domain-specific data and physical AI data.

Lee cited data learning, especially hyper-personalization, as Vogl's biggest differentiator. He said that the more users use voice, messages, and words, the higher the understanding becomes and the more learning continues.

Based on this, Flitto is conducting research and development aimed at expanding Vogl's use across hardware. It is also reviewing ways to combine interpretation and translation solutions with various devices, including portable devices and classroom-only devices.

Lee said he has been considering the device sector for more than two years. He added that the company is currently in the R&D stage and believes it may be able to introduce hardware around 2027.

Flitto plans to strengthen ties between its data business and solution business. It aims to build a structure that applies capabilities accumulated in the data business to its solutions, while also pursuing a structure in which data secured through actual services is fed back to further advance its models.

Lee said that everything starts with the power of data. He added that the company is developing solutions with technology accumulated through its data business, and that a virtuous cycle is forming in the solution development process.

Source: IT DAILY · Yang Seung-gap
Original: https://www.itdaily.kr/news/articleView.html?idxno=241931

References

• https://www.itdaily.kr/news/articleView.html?idxno=241931